Remove obsolete caution about cordoning
As of Kubernetes 1.19, unschedulable nodes are no longer removed from load balancers
This commit is contained in:
@@ -309,13 +309,6 @@ The node controller also adds {{< glossary_tooltip text="taints" term_id="taint"
|
|||||||
corresponding to node problems like node unreachable or not ready. This means
|
corresponding to node problems like node unreachable or not ready. This means
|
||||||
that the scheduler won't place Pods onto unhealthy nodes.
|
that the scheduler won't place Pods onto unhealthy nodes.
|
||||||
|
|
||||||
|
|
||||||
{{< caution >}}
|
|
||||||
`kubectl cordon` marks a node as 'unschedulable', which has the side effect of the service
|
|
||||||
controller removing the node from any LoadBalancer node target lists it was previously
|
|
||||||
eligible for, effectively removing incoming load balancer traffic from the cordoned node(s).
|
|
||||||
{{< /caution >}}
|
|
||||||
|
|
||||||
### Node capacity
|
### Node capacity
|
||||||
|
|
||||||
Node objects track information about the Node's resource capacity: for example, the amount
|
Node objects track information about the Node's resource capacity: for example, the amount
|
||||||
|
|||||||
Reference in New Issue
Block a user